co-author: autonomous vehicles publication

Experience implementing the assessment methodology described herein MTurk demonstrates that this approach can be applied in a cost-effective manner for identifying design concepts that may be appropriate for more detailed development and testing. Since a relatively large number of elements or minor variations can be tested through a series of MTurk runs in a matter of hours (as opposed to weeks or months for focus groups or experimental simulation or field testing), designers should be able to refine concepts both more quickly and in more depth than available development resources typically allow. Factors that are often difficult to explore during design phases (e.g. culture, demographics, prior mental model, etc.) can be factored in early in the process. It is worth noting that there is nothing in this method of early-stage design development that is limited to the messaging application explored in this study; it should be equally applicable to work on other design elements such as interior interface icons, graphics, gages, and other forms of information presentation in automotive, consumer electronics, advertising, and other domains.

Careful and extensive testing vehicle-to-pedestrian communication concepts under real-world conditions and with a broad demographic sampling would seem to be indicated before a design is put into general use due the potentially safety-critical implications of miscommunication. Given the inherent costs of real-world validation testing, efficient methods for early-stage concept assessment are highly desirable for narrowing in on designs that are promising and setting aside those less likely to prove out. Further, early stage methods that make it practical to test a large number of minor design variations increase the probability of elucidating subtle considerations that may lead to optimized implementations.